CN102567957B - Method and system for removing reticulate pattern from image - Google Patents

Abstract

The present invention provides a method and system for removing the texture in an image, which is used to solve the problem in the prior art that the processing effect of removing the texture in the image is not good. The method includes: obtaining the texture feature of the image; establishing an image degradation model; and obtaining the image after removing the texture according to the image degradation model. The technical solution of the present invention utilizes the structural feature of the periodic distribution of the texture, so the texture can be removed better and the integrity of the details of the original image can be kept as much as possible.

Description

Method and system for removing texture in image

technical field

The invention relates to a method and a system for removing net pattern in an image.

Background technique

During the printing process, due to the uneven pressure of the inking roller on the substrate, the roughness of the substrate itself is also different, and the accumulation of ink on the substrate is also different, resulting in the depth of the ink on the substrate. Formed a regular, different shades of inking texture related to the texture of the substrate itself, which can be seen as formed by the periodic appearance of basically the same texture units, which we call reticulation. As shown in the image in box 11 of FIG. 1 , FIG. 1 is an example of an image containing texture in the prior art. When scanning the substrate to obtain the image on the substrate, the texture is also scanned and entered into the acquired image. The texture generally affects the integrity or beauty of the image, and it is usually desired to remove the texture when applying these images. grain, and try to keep the content of the original image from being damaged.

In order to remove the moire in the image, the methods commonly used in actual production at present include: (1) scanning the image with a higher resolution, and then reducing the high-resolution image to obtain an image without moire. Although this method is simple and practical, it is limited by the resolution of the scanning device. (2) Treat the texture as the noise in the image, and use the general image filtering method to smooth or denoise the original image. This type of method cannot keep the details of the image well. (3) Use the periodicity of texture features to de-texture in the frequency domain or wavelet domain. Due to the change of the substrate itself, the distortion generated during scanning, etc., the texture itself is not completely in line with the periodicity, and this method cannot remove the texture well.

Generally speaking, the processing effect of removing the texture in the image in the prior art is not good, and no effective solution has been proposed for this problem so far.

Contents of the invention

The main purpose of the present invention is to provide a method and system for removing the texture in an image, so as to solve the problem in the prior art that the processing effect of removing the texture in the image is not good.

In order to achieve the above object, according to one aspect of the present invention, a method for removing moire in an image is provided.

The method for removing the texture in the image of the present invention includes: acquiring the texture structure feature of the image; establishing an image degradation model; and obtaining the image after the texture is removed according to the image degradation model.

Further, the acquisition of the textured structure features of the image includes: determining the textured unit structure; locating the position of each textured unit in the image and the position of each pixel in the textured unit in the textured unit s position.

Further, the determination of the texture unit structure includes: calculating the autocorrelation image of the image; determining the peak pixel point of the autocorrelation image; selecting any peak pixel point as the first endpoint of the texture unit, and selecting the Any two adjacent pixel points that are not collinear at the first end point are used as the second end point and the third end point of the mesh unit; select the pixel points that are adjacent to the first end point, the second end point, and the third end point As the fourth end point of the textured unit, the quadrilateral unit structure formed by the four endpoints is used as the textured unit structure.

Further, the positioning of the position of each texture unit in the image and the position of each pixel in the texture unit in the texture unit includes: according to the texture unit structure, sequentially traversing all Peak point, with the peak point as the endpoint of the texture unit, determine the position of each texture unit in the image; locate the position of each pixel point in the texture unit in its texture unit.

Further, the establishment of the image degradation model refers to determining the reliability of the color of each pixel in the texture unit as a real color according to the color information of the texture unit.

Further, the reliability that the color of each pixel point is a true color includes: calculating the saturation S _{i(x, y) of any pixel point (x, y} ) in each texture unit i, and calculating each The highest saturation SH _i in the texture unit; add up the ratio of the pixel point saturation S _{i(x, y)} at the same position in all texture units to the highest saturation SH _i in the texture unit, and finally take the mean value as each The reliability of the pixel color as the true color, that is,

Further, before performing the step of obtaining the image after removing the texture according to the image degradation model, it also includes: after locating the position of each texture unit in the image in the image, stabilizing the structure of the texture unit performance evaluation; if there is structural deformation of the textured unit, the original image is updated according to the coordinate transformation to obtain the undistorted image of the textured unit.

Further, according to the image degradation model, obtaining the image after removing the texture includes: weighting according to the reliability of the color of each pixel in the neighborhood of the current pixel as a true color and the distance between each pixel and the current pixel The updated pixel value of the pixel point is obtained, and the updated image is the image after de-screening.

其中P_(x，y)为原图像中像素点颜色为真实色的可靠度，C_(x，y)为原图像中像素的颜色；N_(i，j)为坐标为(i，j)的像素点的邻域；D_(x-i，y-j)为与坐标(x，y)\坐标(i，j)之间距离有关的权重，坐标(x，y)距离坐标(i，j)越远，D_(x-i，y-j)的值越小，归一化系数

Further, the pixel value R _{(i, j)} of the pixel point after the update is:

Among them, P _{(x, y)} is the reliability of the color of the pixel in the original image as the true color, C _{(x, y)} is the color of the pixel in the original image; N _{(i, j)} is the coordinate of (i, j) The neighborhood of pixels; D _{(xi, yj)} is the weight related to the distance between coordinates (x, y)\coordinates (i, j), the farther the coordinates (x, y) are from the coordinates (i, j), The smaller the value of D _{(xi, yj)} , the normalization coefficient

According to another aspect of the present invention, a system for removing moire in an image is provided.

The system for removing texture in an image of the present invention includes: a structure analysis module, used to obtain texture features of an image; a degradation model module, used to establish an image degradation model; a texture processing module, used to base the image degradation model on , to obtain the image after removing the texture.

Further, the structure analysis module is also used to determine the texture unit structure; locate the position of each texture unit in the image and the position of each pixel in the texture unit in the texture unit.

Further, the structure analysis module is also used to calculate the autocorrelation image of the image; determine the peak pixel point of the autocorrelation image; select any peak pixel point as the first endpoint of the mesh unit, and select Any two adjacent pixels of an endpoint non-collinear are used as the second endpoint and the third endpoint of the mesh unit; the pixels adjacent to the first endpoint, the second endpoint, and the third endpoint are selected as For the fourth end point of the textured unit, the quadrilateral unit structure formed by the four endpoints is used as the textured unit structure.

Further, the structure analysis module is also used for traversing all the peak points sequentially according to the texture unit structure, using the peak point as the endpoint of the texture unit to determine the position of each texture unit in the image; The position of each pixel in its texture unit.

Further, the degradation model module is also used to determine the reliability of the color of each pixel in the texture unit as a true color according to the color information of the texture unit.

Further, the degradation model module is also used to calculate the saturation S _{i(x, y) of any pixel point (x, y)} in each texture unit i, and at the same time calculate the highest saturation in each texture unit SH _i ; Accumulate the ratio of the pixel point saturation S _{i(x, y)} at the same position in all textured units to the highest saturation SHi in the textured unit, and finally take the average value as a reliable guarantee that the color of each pixel point is the true color degree, that is

Further, the system also includes a stability module, which is used to evaluate the stability of the texture unit structure according to the position of each texture unit in the image after positioning; if there is a texture deformation of the texture unit , then the original image is updated according to the coordinate transformation to obtain the undistorted image of the mesh unit.

Further, the texture processing module is also used to obtain the updated pixels of the pixel according to the reliability of the color of each pixel in the neighborhood of the current pixel as a true color and the weight of the distance between each pixel and the current pixel. value, the updated image is the image after removing the texture.

计算所述更新后所述像素点的像素值R_(i，j)，其中P_(x，y)为原图像中像素点颜色为真实色的可靠度，C_(x，y)为原图像中像素的颜色；N_(i，j)为坐标为(i，j)的像素点的邻域；D为与距离有关的权重，坐标(x，y)距离坐标(i，j)越远，D的值越小，归一化系数 Further, the texture processing module is also used for

Calculate the pixel value R _{(i, j)} of the pixel point after the update, where P _{(x, y)} is the reliability of the true color of the pixel point color in the original image, and C _{(x, y)} is the reliability in the original image The color of the pixel; N _{(i, j)} is the neighborhood of the pixel with the coordinates (i, j); D is the weight related to the distance, the farther the coordinate (x, y) is from the coordinate (i, j), the D The smaller the value, the normalization coefficient

According to the technical solution of the present invention, the structural feature of the periodic distribution of the texture is utilized, so the texture can be removed better and the integrity of the details of the original image can be kept as much as possible.

Description of drawings

The accompanying drawings in the description are used to provide a further understanding of the present invention and constitute a part of the present application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention. In the attached picture:

Fig. 1 is the example that comprises the image of texture in the prior art;

2 is a schematic diagram of main steps of a method for removing texture in an image according to an embodiment of the present invention;

Fig. 3 is a schematic diagram of a specific step of acquiring the texture feature of an image according to an embodiment of the present invention;

4 is a schematic diagram of a textured cell structure according to an embodiment of the present invention; and

Fig. 5 is a schematic diagram of main modules of a system for removing moire in an image according to an embodiment of the present invention.

Detailed ways

It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and examples.

Fig. 2 is a schematic diagram of the main steps of a method for removing texture in an image according to an embodiment of the present invention. As shown in Fig. 2, the method mainly includes the following steps:

Step S21: Obtain the texture feature of the image;

Step S23: establishing an image degradation model;

Step S25: According to the image degradation model, obtain the image after removing the texture.

The above steps will be further described below. A specific manner of step 21 is shown in FIG. 3 , which is a schematic diagram of a specific step for acquiring texture features of an image according to an embodiment of the present invention. Figure 3 includes the following steps:

Step S31: determining the texture unit structure;

Step S33: Locating the position of each textured unit in the image and the position of each pixel in the textured unit in the textured unit.

In this step, the texture image is analyzed to obtain texture features. Assuming that the image can be divided into different blocks, and the destruction of the image content by the texture in each block obeys the same or similar law, such a block is called a texture unit. Ideally (the texture is regularly arranged regularly, the substrate is not deformed and distorted, etc.), the texture image can be divided into periodically arranged blocks of the same size and shape, and each block is a texture unit. In a non-ideal textured image, different textured units have similar shapes and approximately follow a periodic distribution.

Analysis of texture features Find out the structure and distribution of textured units in the image, such as finding the position of each textured unit in the image, and the relative position of each element in the textured unit in the textured unit Location.

In the above step S31, specifically, the autocorrelation image of the image may be calculated first, then the peak pixel point of the autocorrelation image is determined, and any peak pixel point is selected as the first end point of the mesh unit, and the first endpoint corresponding to the first endpoint is selected. Any two adjacent pixel points whose endpoints are not collinear are used as the second endpoint and the third endpoint of the mesh unit; then select the pixels adjacent to the first endpoint, the second endpoint and the third endpoint as The fourth end point of the textured unit, finally, the quadrilateral unit structure formed by the above four endpoints is used as the textured unit structure. A possible mesh structure is shown in Figure 4, and Figure 4 is a schematic diagram of a texture unit structure according to an embodiment of the present invention, wherein a part of an image is inside a box 41, and a black dot such as a dot 42 represents in the image The center of the mesh unit, the quadrilateral surrounding each dot, such as the quadrilateral 43 is a mesh unit.

In the above-mentioned step S33, specifically, it may be first to traverse all the peak points sequentially according to the texture unit structure, and then use the peak point as the endpoint of the texture unit to determine the position of each texture unit in the image; The position of the pixel in its texture unit.

Affected by the texture, the color of the pixel will change, and pixels of different positions and colors in the texture unit are affected by the texture to a different degree (for example, the saturation of pixels that are greatly affected by the texture in the texture image is often low) , different degrees of color changes will occur, and the greater the degree of change, the smaller the probability of being the same as the real color. When establishing the image degradation model in step S23, it may be based on the color information of the texture unit to determine the reliability of the color of each pixel in the texture unit being a true color. Specifically, the saturation S _{i(x, y) of any pixel point (x, y} ) in each texture unit i can be calculated first, and the highest saturation SH _i in each texture unit can be calculated at the same time; The ratio of the pixel point saturation S _{i (x, y)} at the same position in the grain unit to the highest saturation SH _i in the texture unit, and finally take the mean value as the reliability P _{(x, y)} of the true color of each pixel point color _y) , ie

Before the above-mentioned step S25, the structure of the texture unit can be evaluated for stability according to the position of each texture unit in the image in the positioning image; The original image is the undistorted image of the textured unit.

其中C_(x，y)为原图像中像素的颜色；N_(i，j)为(i，j)的邻域；D_(x-i，y-j)为与坐标(x，y)\坐标(i，j)之间距离有关的权重，坐标(x，y)距离坐标(i，j)越远，D_(x-i，y-j)的值越小，归一化系数 Taking the probability-based image degradation model obtained above as an example, for each textured unit in the textured image, the color reliability of a certain pixel can be obtained by searching the meshed unit with a standard shape (in non-ideal textured In some cases, it is necessary to obtain its coordinates in the standard mesh unit through the transformation of the coordinates). A possible image restoration method is, in the above-mentioned step S25, according to the reliability of the color of each pixel in the neighborhood of the current pixel as the real color and the weight of the distance between each pixel and the current pixel to obtain the updated pixel The pixel value of , the updated image is the image after removing the texture. The pixel value R _{(i, j)} of the updated pixel point can be calculated by the following formula:

Among them, C _{(x, y)} is the color of the pixel in the original image; N _{(i, j)} is the neighborhood of (i, j); D _{(xi, yj)} is the coordinate (x, y)\coordinate (i, The weight related to the distance between j), the farther the coordinate (x, y) is from the coordinate (i, j), the smaller the value of D _{(xi, yj)} , and the normalization coefficient

Fig. 5 is a schematic diagram of main modules of a system for removing moire in an image according to an embodiment of the present invention. As shown in Figure 5, the system 50 for removing the texture in the image mainly includes the following modules: a structure analysis module, which is used to obtain the texture characteristics of the image; a degradation model module, which is used to establish an image degradation model; a texture processing module, which uses According to the image degradation model, the image after removing the texture is obtained.

The structure analysis module can also be used to determine the texture unit structure; locate the position of each texture unit in the image and the position of each pixel in the texture unit in the texture unit.

The structural analysis module can also be used to calculate the autocorrelation image of the image; determine the peak pixel point of the autocorrelation image; select any peak pixel point as the first endpoint of the mesh unit, and select any two non-collinear with the first endpoint adjacent pixel points, as the second end point and the third end point of the mesh unit; select the pixels adjacent to the first end point, the second end point and the third end point as the fourth end point of the mesh unit, with the above-mentioned A quadrilateral unit structure composed of four endpoints is used as a textured unit structure.

The structural analysis module can also be used to traverse all the peak points in turn according to the structure of the texture unit, and use the peak point as the endpoint of the texture unit to determine the position of each texture unit in the image; locate each pixel in the texture unit in its The position in the texture unit. The degradation model module can also be used to determine the reliability of the color of each pixel in the texture unit as the real color according to the color information of the texture unit.

The degradation model module can also be used to calculate the saturation S _{i(x, y) of any pixel point (x, y)} in each texture unit i, and simultaneously calculate the highest saturation SH _i in each texture unit; accumulate all The ratio of the pixel saturation S _{i(x, y)} at the same position in the texture unit to the highest saturation SH _i in the texture unit, and finally take the mean value as the reliability of the color of each pixel point as the real color, that is

The system 50 shown in FIG. 5 can also include a stability module (not shown in the figure), which is used to evaluate the stability of the texture unit structure according to the position of each texture unit in the image in the positioning image; If there is a structural deformation of the textured unit, the original image is updated according to the coordinate transformation to obtain an undistorted image of the textured unit.

The mesh processing module can also be used to obtain the updated pixel value of the pixel according to the reliability of the true color of each pixel in the neighborhood of the current pixel and the distance between each pixel and the current pixel. The image is the image after removing the texture.

计算更新后像素点的像素值R_(i，j)，其中P_(x，y)为原图像中像素点颜色为真实色的可靠度，C_(x，y)为原图像中像素的颜色；N_(i，j)为(i，j)的邻域；D_(x-i，y-j)为与坐标(x，y)\坐标(i，j)之间距离有关的权重，坐标(x，y)距离坐标(i，j)越远，D_(x-i，y-j)的值越小，归一化系数 Texture processing module can also be used according to the

Calculate the pixel value R _{(i, j)} of the updated pixel point, where P _{(x, y)} is the reliability of the true color of the pixel point color in the original image, and C _{(x, y)} is the color of the pixel in the original image; N _{(i, j)} is the neighborhood of (i, j); D _{(xi, yj)} is the weight related to the distance between coordinates (x, y)\coordinates (i, j), coordinates (x, y) The farther away from coordinates (i, j), the smaller the value of D _{(xi, yj)} , the normalization coefficient

It can be seen from the above description that, compared with the general image restoration method, the technical solution of the embodiment of the present invention utilizes the structural characteristics of the periodic distribution of the texture, so it can better remove the texture and keep the texture as much as possible. Integrity of original image detail.

Obviously, those skilled in the art should understand that each module or each step of the above-mentioned present invention can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network formed by multiple computing devices Optionally, they can be implemented with program codes executable by a computing device, so that they can be stored in a storage device and executed by a computing device, or they can be made into individual integrated circuit modules, or they can be integrated into Multiple modules or steps are fabricated into a single integrated circuit module to realize. As such, the present invention is not limited to any specific combination of hardware and software.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (14)

1. A method for removing texture in an image, characterized in that, comprising, Obtain the texture feature of the image; Build an image degradation model; According to the image degradation model, the image after removing the texture is obtained, Wherein, the texture feature of the acquired image comprises: Determine the texture unit structure; Positioning the position of each textured unit in the image and the position of each pixel in the textured unit in the textured unit, Wherein, said determining the texture unit structure includes: Calculate the autocorrelation image of the image; Determine the peak pixel point of the autocorrelation image; Selecting any peak pixel point as the first endpoint of the mesh unit, and selecting any two adjacent pixel points that are non-collinear with the first endpoint as the second endpoint and the third endpoint of the texture unit; Select the pixel points adjacent to the first end point, the second end point and the third end point as the fourth end point of the mesh unit End points, the quadrilateral unit structure formed by the four end points is used as the mesh unit structure. 2. The method according to claim 1, wherein the position of each textured unit in the image and the position of each pixel in the textured unit in the textured unit in the positioning image comprise : According to the texture unit structure, traverse all the peak points in turn, and use the peak point as the endpoint of the texture unit to determine the position of each texture unit in the image; Locate the position of each pixel in the texture unit in its texture unit. 3. The method according to claim 1, wherein said establishing the image degradation model refers to determining the reliability of the color of each pixel in the texture unit as a true color according to the color information of the texture unit. 4. method according to claim 3, is characterized in that, the reliability that described each pixel point color is true color comprises: Calculate the saturation S _{i(x,y) of any pixel point (x,y)} in each texture unit i, and calculate the highest saturation SH _i in each texture unit; According to the following formula, add up the ratio of the pixel saturation S _{i(x, y)} at the same position in all textured units to the highest saturation SH _i in the textured unit, and finally take the average as the color of each pixel as the true color The reliability P _(x,y) of , the formula is: $P_{(x,\mathrm{the\; y})}=\frac{1}{\mathrm{no}}\underset{i=1...\mathrm{no}}{\mathrm{\Σ}}\frac{S_{i(x,\mathrm{the\; y})}}{{\mathrm{SH}}_i}.$ 5. The method according to any one of claims 1 to 4, characterized in that, before performing the step of obtaining the image after removing the texture according to the image degradation model, further comprising: according to each The position of the texture unit in the image is used to evaluate the stability of the structure of the texture unit; if there is a deformation of the texture unit structure, the original image is updated according to the coordinate transformation to obtain the undeformed image of the texture unit. 6. The method according to claim 3 or 4, wherein, according to the image degradation model, obtaining the image after removing the texture comprises: According to the reliability of the true color of each pixel in the neighborhood of the current pixel and the weighted distance between each pixel and the current pixel, the pixel value of the updated pixel is obtained, and the updated image is the image after removing the texture. Image. 7. The method according to claim 6, wherein the pixel value R _{(i, j)} of the pixel point after the update is: P _(x,y) D _(xi,yj) C _(x,y) , where P _(x,y) is the reliability of the pixel color in the original image as the real color, and C _(x,y) is the reliability of the original image The color of the pixel; N _(i,j) is the neighborhood of the pixel with coordinates (i,j); D _(xi,yj) is related to the distance between coordinates (x,y)\coordinates (i,j) The weight of the coordinates (x, y) is farther away from the coordinates (i, j), the smaller the value of D _{(xi, yj)} , the normalization coefficient

P _(x,y) D _(xi,yj) . 8. A system for removing texture in an image, comprising: Structural analysis module, used to obtain the texture feature of the image; The degradation model module is used to establish an image degradation model; A texture processing module, configured to obtain a texture-removed image according to the image degradation model, Wherein, the structure analysis module is also used to determine the texture unit structure; locate the position of each texture unit in the image and the position of each pixel in the texture unit in the texture unit, Wherein, the structure analysis module is also used to calculate the autocorrelation image of the image; determine the peak pixel point of the autocorrelation image; select any peak pixel point as the first endpoint of the mesh unit, and select the Any two adjacent pixel points whose endpoints are not collinear are used as the second endpoint and the third endpoint of the mesh unit; the pixels adjacent to the first endpoint, the second endpoint and the third endpoint are selected as the mesh The fourth end point of the grain unit, the quadrilateral unit structure formed by the four end points is used as the texture unit structure. 9. The system according to claim 8, wherein the structure analysis module is further used to traverse all peak points sequentially according to the structure of the texture unit, and use the peak point as the endpoint of the texture unit to determine each The position of each texture unit; locate the position of each pixel in the texture unit in its texture unit. 10 . The system according to claim 8 , wherein the degradation model module is further configured to determine the reliability that the color of each pixel in the texture unit is a true color according to the color information of the texture unit. 11 . 11. The system according to claim 10, wherein the degradation model module is also used to calculate the saturation S _{i(x, y) of any pixel point (x,} y) in each texture unit i , and calculate the highest saturation SH _i in each texture unit at the same time; accumulate the ratio of the pixel saturation S _i(x,y) at the same position in all texture units to the highest saturation SH _i in the texture unit, and finally Take the mean value as the reliability P _(x,y) that the color of each pixel point is the true color, that is

12. The system according to any one of claims 8 to 11, characterized in that, it also includes a stability module, which is used for adjusting the position of each texture unit in the image according to the position of each texture unit in the image The stability of the structure is evaluated; if there is structural deformation of the textured unit, the original image is updated according to the coordinate transformation to obtain the undeformed image of the textured unit. 13. The system according to claim 10 or 11, characterized in that, the texture processing module is also used for the reliability of the color of each pixel in the neighborhood of the current pixel as a true color and the relationship between each pixel and the current color. The pixel value of the pixel point after updating is obtained by weighting the distance of the pixel point, and the updated image is the image after removing the texture. 14. The system according to claim 13, wherein the anilox processing module is also used for

P _(x,y) D _(xi,yj) C _(x,y) calculates the pixel value R _(i,j) of the pixel after the update, where P _(x,y) is the pixel in the original image The color is the reliability of the real color, C _{(x, y)} is the color of the pixel in the original image; N _{(i, j)} is the neighborhood of the pixel whose coordinates are (i, j); D _{(xi, yj)} is The weight related to the distance between coordinates (x, y)\coordinates (i, j), the farther the coordinates (x, y) from the coordinates (i, j), the smaller the value of D _{(xi, yj)} , normalized Coefficient $K_{(i,j)}=\underset{x,\mathrm{the\; y}\∈N_{(i,j)}}{\mathrm{\Σ}}$ P _(x,y) D _(xi,yj) .

Images